Hindawi BioMed Research International Volume 2018, Article ID 7456894, 12 pages https://doi.org/10.1155/2018/7456894

Research Article The Effect of Breed, Gender, and Acid Stimulation in Saliva Proteome

Sónia Lucena,1,2 Ana V. Coelho ,3 Fernando Capela-Silva ,1,4 Asta Tvarijonaviciute ,5 and Elsa Lamy 1

1 Instituto de Cienciasˆ Agrarias´ e Ambientais Mediterranicasˆ (ICAAM), Universidade de Evora,´ 7000-083 Evora,´ Portugal 2Departamento de Medicina Veterinaria,EscoladeCi´ enciasˆ e Tecnologia, Universidade de Evora,´ 7000-083 Evora,´ Portugal 3Instituto de Tecnologia Qu´ımica e Biologica´ Antonio´ Xavier, Universidade Nova de Lisboa, 2780-157 Oeiras, Portugal 4Departamento de Biologia, Escola de Cienciasˆ e Tecnologia, Universidade de Evora,´ 7000-671 Evora,´ Portugal 5Interdisciplinary Laboratory of Clinical Analysis (Interlab-UMU), Regional Campus of International Excellence “Campus Mare Nostrum”,UniversityofMurcia,30100Espinardo,Murcia,Spain

Correspondence should be addressed to Elsa Lamy; [email protected]

Received 10 February 2018; Accepted 4 April 2018; Published 3 June 2018

Academic Editor: Giuseppe Piccione

Copyright © 2018 Sonia´ Lucena et al. Tis is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Saliva gained interest as a potential noninvasive source of biomarkers in humans and that interest starts to be extended also to other animal species. For this purpose, the knowledge of the salivary proteome in healthy conditions and the factors that afect it and howtheyafectitarenecessary.Teaimofthepresentstudywastoassesstheefectthatgenderandbreedhaveinsalivaproteome and the changes in it induced by stimulation with acid. Saliva from 4 diferent purebred (, , Rafeiro Alentejano, and ) of both genders was collected without and afer stimulation with lemon juice. SDS-PAGE and two- dimensional gel electrophoresis (2-DE) profles were compared and the proteins of interest in-gel digested and identifed by mass spectrometry. Acid stimulation decreased total protein concentration and the relative amounts of some protein bands/spots. Gender appeared to have minimal efect in saliva proteome, whereas the infuence of breed varies. and Portuguese Podengos were the two breeds with higher diferences. In conclusion, stimulation procedures and should be considered in data analysis when using salivary proteins for diagnostic purposes.

1. Introduction mRNA determination [9]. Furthermore, recently, healthy dog saliva proteome has been characterized by shotgun Physiologicalvariablesareofaddedvaluetoassessthewelfare proteomics, with the identifcation of 2,491 proteins and and lifespan both in humans and in animals, as they pro- peptides [10]. Despite this characterization, two-dimensional vide important information for interpreting and validating electrophoresis (2-DE) salivary protein profles of dog saliva emotional and biological responses, respectively [1]. Saliva have been less explored. Although several researchers con- has gained interest for biomarker identifcation, mainly due to the noninvasive nature of its collection; at the same time sider that gel-based approaches provide limited information, that it contains glandular and blood-born molecules that 2-DE continues providing reliable quantitative results on can change under diferent conditions [2]. In dogs, most of diferential protein expressions as they display a high num- the studies have been focused on the evaluation of stress ber of protein species, their isoforms, and posttranslational bymeasuringsalivarycortisollevels[3].Infectiousagents, modifcations at the same time [11]. It also has the advantage such as Helicobacter spp., Bartonella spp., or rabies virus, of allowing modifcations of the protein mixtures caused by have also been evaluated [4–6]. In addition, canine saliva has inadequate treatment or endogenous protease activities with been used for quantifcation of acute phase proteins [7] and physiological relevance to be easily recognized via pattern allergen measurements [8] and in forensic studies for canine disturbances by 2D gels [11]. 2 BioMed Research International

Table 1: Dog population for each breed by gender and age.

Gender Breed Average body weight (Kg) Age (years) Female Male Portuguese Podengo 4-5 0.5–10 7 6 Greyhound 26–40 1–7 9 6 Rafeiro Alentejano 35–50 0.5–8 8 7 Beagle 9–11 2–11 0 10

In humans, physiological and environmental factors, such sample were collected in all animals, in diferent days. In as gender, age, interindividual variability, taste stimulation, one of these sample collections two to three drops of lemon and circadian rhythms, were identifed to cause diferences juice were put under the tongue for stimulating saliva fow in the human salivary protein profles [12]. However, to the (acid stimulation). Only for Rafeiro Alentejano breed acid best of the author’s knowledge, in dogs such infuences in stimulated saliva collection was not possible. Afer collection, salivary proteome are not deeply studied. Te knowledge the cotton cylinders were immediately placed on ice, until of the possible salivary proteome changes due to diferent laboratory arrival, which lasted no more than 30 minutes. factors would later permit correcting data interpretation for In the laboratory saliva was extracted from the cotton ∘ disease diagnostics. roll by centrifugation at 4 C, at 5000 rpm, for 5 min, and ∘ Diferent methods of saliva sampling in dogs have been immediately stored at −20 C for further analysis. reported in literature: (1) without stimulation [10, 13]; (2) using diferent stimulating methods, such as citric acid in 2.4. Total Protein Concentration. Bradford method protein swabs [14] or in crystals spreader in the tongue [15], beef- assay [22] with BSA as the standard protein (Pierce Biotech- favoured cotton ropes [16], dogs’ snack held in front of the nology, Rockford, IL, USA) was performed to determine dog’s snout [17], or visualization and smell of food [18] what the total protein concentration of each sample. Standards could result in diferent salivary proteomes. Acid stimulation, andsampleswererunintriplicate,in96-wellmicroplates. which is one of the mostly used methods for stimulating saliva Absorbance was read at 600 nm in a microplate reader production in humans, has been already reported to infuence (Glomax, Promega). human salivary proteome [12]. However, its infuence in dog saliva composition has not been reported. 2.5. SDS-PAGE. Proteins from individual saliva samples of Te aims of this study were to evaluate the possible allanimals(bothwithoutandwithacidstimulation)were infuence of biological factors, namely, breed and gender, and separated by SDS-PAGE electrophoresis in 14% acrylamide diferent saliva sampling conditions (with and without saliva gels in a mini-protean apparatus (BioRad) as described before stimulation with citric acid) on dog’s saliva proteome. [23]. Briefy, a total of 15 �g protein from each saliva sample was run in each lane. Te samples were resuspended in 2. Materials and Methods sample bufer [Tris–HCl 0.125 M pH 6.8, 2% (w/v) SDS, 5% (v/v) 2-mercaptoetanol, 20% (v/v) glycerol traces of ∘ 2.1. Ethical Note. Dogs used in this study belong to three bromophenol blue], heated at 95 Cfor5minutes,andrun kennels and to a university (University of Murcia), whose at a constant voltage of 140 V until the dye front reaches gavetheirinformedconsentandparticipatedinthecollection the end of the gel. Gels were fxed in 40% methanol, 20% procedures by handling the animals. Te saliva collection acetic acid, for one hour, stained with Coomassie Brilliant and all animal procedures were carried out by researcher Blue (CBB) G-250 (0.125% CBB G-250, 20% ethanol) for accredited by the Federation of European Laboratory Animal two hours and destained in several washes with distilled Science Association (FELASA) and conformed to legislation. water. A scanning Molecular Dynamics densitometer with internal calibration and LabScan sofware (GE, Healthcare) 2.2. Dog Population. Dog population for each breed by were used to acquire gel images and to determine the per- gender and age is shown in Table 1. All were healthy and centageofvolumeofeachproteinband;GelAnalyzersofware normal weight animals. Only male’s pure breed Beagle were (http://www.gelanalyzer.com/) was used to analyze the gel neutered animals. images. Molecular masses were determined in accordance with molecular mass standards (Bio-Rad Precision Plus 2.3. Saliva Collection. Saliva samples were collected in the Protein Dual Color 161–0394) run with protein samples. afernoon between 3:30 and 6:30 pm. Dogs did not eat for 16–18 hours prior to saliva sampling. Water was provided 2.6. Two-Dimensional Gel Electrophoresis (2-DE) ad libitum. Saliva was collected by rolling a cotton cylinder (Salivette5, Sarstedt) inside each dog’s mouth as described 2.6.1. Protein Precipitation. Due to the limited amount of previously [19, 20]. Te cotton cylinders were inserted under individual saliva samples, the unstimulated and acid stimu- the dog’s tongue for chew, until completely soaked with lated saliva samples from each breed and gender were mixed saliva, for a maximum of two minutes [21]. Two to three in pools, constituting a total of 12 pools: (1) unstimulated BioMed Research International 3 female Portuguese Podengo; (2) unstimulated male Por- Peptide mass spectra were acquired using a MALDI- tuguese Podengo; (3) stimulated female Portuguese Podengo; TOF/TOF 4800 plus MS/MS (Applied Biosystems5 Life (4) stimulated male Portuguese Podengo; (5) unstimulated Technologies, Carlsbad, United States of America). Data femaleGreyhound;(6)unstimulatedmaleGreyhound;(7) were acquired in positive MS refector using a PepMix1 stimulatedfemaleGreyhound;(8)stimulatedmaleGrey- (LaserBio Labs, Sophia-Antipolis, France) to calibrate the ; (9) unstimulated female Rafeiro Alentejano; (10) instrument. Each refector MS spectrum was collected in a unstimulated male Rafeiro Alentejano; (11) unstimulated result independent acquisition mode, using 750 shots per male Beagle; (12) stimulated male Beagle. Volumes of saliva spectra in 800–4000 �/� range and fxed laser intensity from each pool containing 250 �g of total protein were used. to 3100 V. Fifeen of the strongest precursors were selected Te volume of each pool was mixed with equal volume of for MS/MS. MS/MS analyses were performed using CID ∘ TCA 20% (m/v), incubated overnight, at −20 C, followed (Collision Induced Dissociation) assisted with a collision by centrifugation at 15,000�, 30 min, and two cold-acetone energy of 1 kV and a gas pressure of 1 × 10−6 Torr. For each washes. Tis protocol as previously observed by us allows MS/MS spectrum, 1400 laser shots were collected, using fxed satisfactory results for preparation of dog saliva samples for laser intensity of 4400 V.Processing and interpretation of MS 2-DE [24]. andMS/MSspectrawereperformedwiththe4000Series Explored6 Sofware(AppliedBiosystems5 Life Technologies, 2.6.2. 2-DE Protein Separation. For 2-DE, the precipitates Carlsbad, United States of America). were mixed with 250 �Lrehydrationbufer[7Murea,2M Protein identifcation was performed using MS and thiourea, 4% (w/v) CHAPS, 2% (v/v), 60 mM DTT and MS/MS spectral data and ProteinPilot (Applied Biosys- traces of bromophenol blue] +5 �LIPGbufer+5�LNaOH. tems,version3.0,rev.114732)onCaniscanisdatabase Ten the precipitates were sonicated until total resuspension (85118 sequences; 46,697,962 residues) retrieved from NCBI and incubated during 1 h at room temperature, being subse- (downloaded in October 2017). Searches included trypsin quently centrifuged for 5 min at 10000 rpm. IPG strips (13 cm, as digesting enzyme; peptide mass tolerance of 50 ppm; pH 3–10 NL; GE, Healthcare) were passively rehydrated fragment mass tolerance of 0.5 Da and possible oxidation, overnight with this solution. Focusing was performed in a ∘ carbomidomethylation, or deaminidation as variable amino Multiphor II (GE, Healthcare) at 20 C, with the programme acid modifcations with one missed cleavage. Peptides were (gradient): (1) 0–300 V for 2 h; (2) 300 V for 2 h; 300 V only considered if the ion score indicated extensive homology to 3500V for 6h; 3500V for 6h. Focused strips were (� < 0.05). Proteins were considered if the protein score equilibrated in two steps of 15 min each with equilibration indicated signifcant statistical confdence (� < 0.05). bufer [50 mM Tris–HCl, pH 8.8; 6 M urea; 30% (v/v) glycerol Protein identifcations with only one matched peptide were and 2% (w/v) SDS], with the addition of 1% (w/v) DTT considered if they were identifed with >95% confdence. and 65 mM iodoacetamide in the frst and second steps, respectively. Afer equilibration the strips were applied in 2.8. Statistical Analysis. Multivariate analyse of protein the top of a SDS-PAGE gel 14% acrylamide and run at bands, on one hand, and protein spots, on the other, were 150 V constant voltage in a mini-protean system (BioRad). performed with MetaboAnalyst 3.6 to evaluate clustering Staining with CBB-G250 and destaining were done through of individuals or groups [26]. Data normalization was used the same protocol described for SDS-PAGE gels. Gel images when normal distribution was not observed, using trans- were acquired using the same scan method and apparatus formation (log10) or scaling methods, alone or combined. described for SDS-PAGE gels. ImageMaster 2D Platinum Te method chosen was the one that allowed data to be v7 sofware was used to analyze these gel images. Spot normally distributed. For univariate analysis, �-test, one-way editing and the match were performed automatically and ANOVA, and two-way ANOVA were used for comparison of corrected manually. Spot volume was normalized to the total protein profles (band percentage volume or spots percentage spot volume. Tree laboratorial replicates of each pool were volume)betweenunstimulatedandacidstimulatedsalivaand run. among breeds and genders. For Multivariate Analysis, par- tial least squares discriminant analysis (PLS-DA) was used. 2.7. Protein Identifcation. Bands and spots that difered Discriminant variables selection was done using variable among the factors tested were manually excised from importance in the projection (VIP) with a threshold of 1.0. gels and digested with trypsin following the protocol Finally, paired-samples �-test was used for comparison of already described [25]. MALDI-TOF/-TOF mass spectrom- total protein concentration between saliva samples with and etry was used for protein identifcations. Tryptic peptide without stimulation. Statistical signifcance was considered mixtures were acidifed with 5% (V/V) formic acid, desalted, for � < 0.05. and concentrated using home-made reversal phase (R2 pores-Applied Biosystems) microcolumns (R2 pores-Applied 3. Results Biosystems). Peptides were eluted with the matrix solution � ( -cyano-4-hydroxycinnamic acid Fluka) 5 mg/mL in 50% 3.1. Efect of Acid Stimulation on Salivary Proteome (v/v) acetonitrile and 5% (v/v) formic acid. MS and MS/MS data were acquired in positive refector mode in a 4800 Plus 3.1.1. Total Protein Concentration. Total protein concentra- AB SCIEX using the sofware 4000 Series Explorer, version tion decreased signifcantly in stimulated saliva in males of 3.5.3.3 (Applied Biosystems). both pure breeds Portuguese Podengo and Beagle. In females, 4 BioMed Research International

Table 2: Comparison of total protein concentration (mean ± standard error) between saliva with acid stimulation and saliva without stimulation, for each dogs breed and gender.

Total protein concentration (�g/mL) with acid stimulation without stimulation � Breed ∗ Portuguese Pondego (� = 6) 843.0 ± 163.6 2385.7 ± 482.9 0.036 Greyhound (� = 7) 961.7 ± 72.3 1146.7 ± 504.7 0.354 ∗ Beagle (� = 7) 1273.3 ± 161.8 1811.8 ± 246.3 0.033 Gender Female (Podengo, �=4,Greyhound,�=4) 950.1 ± 115.3 1743.3 ± 404.3 0.112 ∗ Male (Beagle, �=7,Podengo,�=2,Greyhound,�=3)1049.2± 110.5 1737.3 ± 170.7 0.001 ∗ Statistically signifcant diferences for � < 0.05.

Rw/o Rw/o Rw/o Gw/o Gw/o Gw/o Pw/o Pw/o MW kDa Bw/o Bw Bw/o Bw

250 A 150 B 100 C 75 D E F 50 G 37

I J 25 K 20

15 M

10

(a) (b)

Figure 1: Representative SDS-PAGE profle of dog saliva: (a) from diferent breeds without stimulation (R: Rafeiro; G: Greyhound; P: Portuguese Podengo); (b) from Beagles without (w/o) and with (w) acid stimulation; MW: molecular mass marker; upper letters indicate the diferent protein bands.

no statistically signifcant diferences were observed for saliva in their intensities/volumes, which were induced by acid collected under the two conditions (Table 2). Concerning stimulation. Some of these changes were observed to be salivary fow rate, although this was not measured, it was dependent on the dogs’ breed and/or gender. Considering possible to observe a tendency for higher salivary fow rates the total of the animals, 2 of the protein bands decreased (F in big, comparatively to small breeds and higher salivary fow and J) and one increased (I1) with acid stimulation (Table 3). rate afer lemon juice induction, in all breeds. Concerning bands F and J, the decreased levels were observed only in males and not in females. 3.1.2. SDS-PAGE Profle. Among the 16 protein bands, with By considering the dog breeds separately, changes molecular masses between 20 and 245 kDa, observed in SDS- induced by stimulation were observed only for Beagles: PAGE protein profles (Figure 1), some presented changes decreased expression levels of 4 protein bands (B, D, F, and J) BioMed Research International 5

Table 3: Protein bands diferently expressed (mean ± standard error) between saliva collected with and without acid stimulation.

%vol Bands � Without acid stimulation With acid stimulation Total of animals (� = 20) ∗ F 8.26 ± 0.46 5.61 ± 0.58 0.002 ∗ I1 3.69 ± 0.51 7.62 ± 1.01 0.002 ∗ J12.81± 0.54 9.24 ± 0.63 0.0008 Beagles (only males) (� = 7) ∗ B9.10± 0.79 6.13 ± 0.39 0.004 ∗ D 10.28 ± 1.04 6.9 ± 0.39 0.004 ∗ F 8.34 ± 0.91 4.15 ± 0.64 0.002 ∗ I1 3.79 ± 1.00 10.10 ± 1.6 0.010 ∗ J 11.88 ± 0.98 8.43 ± 0.70 0.002 Males (three breeds) (� = 12) ∗ F 8.34 ± 0.54 5.29 ± 0.69 0.003 ∗ J 12.88 ± 0.68 8.56 ± 0.79 0.0003 ∗ Statistically signifcant diferences for � < 0.05.

Table 4: Mass spectrometry identifcation of proteins present in bands from saliva SDS-PAGE profles.

Matched NCBI Accession Code Estim/theoret ∗ Seq. Cov. Band Protein # ID Score Peptides Accession n MW (kDa) (%) MS (MS/MS) A Mucin-19 XP 022267206.1 240.6/340.8 201 11 21 (5) C IgGFc-binding protein XP 022261796.1 75/318.0 187 14 14 (9) Chain A, Crystal Structure Analysis D pdb|5GHK|A 67.8/65.7 815 52 15 (11) Of Canine Serum Albumin E Serum albumin isoform X1 XP 005628024.1 61.3/68.6 661 44 12 (10) F IgGFc-binding protein XP 022261796.1 52.6/318.0 313 8 13 (6) Full-double-headed protease M sp|P01002.1|IPSG CANLF 12.2/12.8 166 46 6 (3) inhibitor, submandibular gland # ∗ MW values observed in gel versus theoretical ones. Protein score is −10∗log(�),where� is the probability that the observed match is a random event. Protein scores greater than 62 are signifcant (� < 0.05). and increased expression level of 1 protein band (I1) (Table 3). observed to be present in lower volume in the saliva collected Information about mass spectrometry details of identifed afer stimulation: spot 0 (34.4 ± 5.14 and 13.9 ± 2.97% vol., proteins is present in Table 4. saliva without and with stimulation, respectively), spot 5 Although, in the pure breeds Portuguese Podengo and (0.73 ± 0.02 and 0.47 ± 0.05% vol., saliva without and with Greyhound, none of the individual bands from SDS-PAGE stimulation, respectively) and spot 81 (0.5 ± 0.24 and 0.21 ± protein profles showed statistical signifcant intensity dif- 0.20% vol., saliva without and with stimulation). Tese spots ferences, between the saliva collected with and without were identifed as serum albumin subunit A, cytoskeletal acid stimulation, the multivariate PLS-DA model clustered keratin, and one unknown protein (Table 5). separately unstimulated saliva from acid stimulated saliva, in these two breeds (Figure 2). Te protein bands J, K, 3.2. Efect of Dog’s Breeds and Genders on Salivary Proteome andMwerethemajorcontributorsforthediferencesin . Band M was identifed as containing full- 3.2.1. Total Protein Concentration. Te four diferent breeds double-headed protease inhibitor, whereas the other two did not difer among them for the total protein concentration bands resulted in no confdent identifcation. Te protein of saliva, as shown by univariate statistical analysis. Also, no bands C, E, and G, identifed as containing IgGFc-binding diferences were observed between genders, neither for saliva protein and serum albumin, were the major contributors for collected without nor saliva collected with acid stimulation. diferences in Portuguese Podengos (Supplementary Figure 1). 3.2.2. SDS-PAGE Profle. Salivary protein profles of the 4 dog breeds studied were compared for the saliva collected without 3.1.3. Two-Dimensional Protein Profle (2-DE). By analyzing acid stimulation. Six protein bands showed a diferent vol- 2-DE salivary protein profles (Figure 3), 3 protein spots were ume among dog breeds (Table 6): bands containing serum 6 BioMed Research International

Scores Plot Scores Plot

10 10

5 5 %) 10.8

0 0 Component 2 (13.2%) Component Component 2 ( Component −5 −5

−10 −10 −20 −15 −10 −5 0510 15 −20 −10 0 10 20 Component 1 (24.9%) Component 1 (45%) 0 0 1 1 (a) (b) Scores Plot Scores Plot 15 1.5

10 1.0

5 0.5 %) %) 14

20.7 0.0 0

−0.5 −5 Component 2 ( Component Component 2 ( Component −1.0 −10

−1.5 −15

−2.0 −15 −10 −5 0 510 −1.5 −1.0 −0.5 0.0 0.5 1.0 1.5 2.0 Component 1 (22.4%) Component 1 (34%) 0 0 1 1 (c) (d)

Figure 2: PLS-DA of saliva samples SDS-PAGE bands for all dogs (a), Beagles (� = 7) (b), Greyhounds (� = 7) (c), and Portuguese Podengos (� = 6) (d). Scaling was applied to rows when needed; � and � axes show principal component 1 (PC1) and principal component 2 (PC2), respectively, and the total variance explained by each of them. Δ: with acid stimulation; +: without stimulation. albumin were observed to be increased in Beagles, whereas a trends for gender were found and no relationship between band containing a full-double-headed protease inhibitor was breed and gender was found, as well. decreased, comparatively to the other breeds; bands contain- Trough the multivariate PLS-DA model, that has into ing albumin and IgGFc-binding protein were increased and account the interrelationship among variables, it was possible one not identifed was decreased in Portuguese Podengo. No to cluster Portuguese Podengos and Beagles more distant, BioMed Research International 7

Table 5: Mass spectrometry identifcation of proteins present in spots from saliva 2-DE profles difering between stimulation conditions and/or among breeds.

Estim/ Matched Entry Estim/ Score %Seq. Spots Protein theoret ∗ Peptides reference theor pI ID Cov. MW (kDa) MS (MS/MS) Chain A, Crystal Structure Analysis Of 0 pdb|5GHK|A (NCBI) 78.1/65.7 4.9/5.3 263 42 18 (2) Canine Serum Albumin 5 Keratin, type I cytoskeletal 10 Q6EIZ0 (Uniprot) 18.5/57.7 7.8/5.1 207 30 10 (5) double-headed protease inhibitor, 8 XP 022264993.1 (NCBI) 17.9/15.7 6.0/8.6 428 58 4 (8) submandibular gland 12 Immunoglobulin J chain XP 532398.2 (NCBI) 30.1/18.3 4.4/4.7 125 40 3 (2) Immunoglobulin lambda-1 light chain 16 XP 005636600.1 (NCBI) 30.0/24.8 6.0/6.4 326 33 6 (4) isoform X34 Immunoglobulin lambda-1 light chain 18 XP 022266294.1 (NCBI) 31.0/24.9 5.5/5.1 198 35 8 (4) isoform X25 37 IgGFc-binding protein XP 022261796.1 (NCBI) 59.9/318.0 4.9/5.2 267 4 7 (4) 45 Uncharacterized protein J9P732 (Uniprot) 25.0/21.4 5.8/6.0 192 28 4 (5) 81 Uncharacterized protein F1PW98 (Uniprot) 19.1/55.0 8.0/5.7 111 29 15 (2) ∗ Protein score is −10∗log(�),where� is the probability that the observed match is a random event. Protein scores greater than 62 are signifcant (� < .05).

3 10 MW (kDa) of immunoglobulin lambda-1). Spots 8 (� = 0.043) and 26 150 (� = 0.015) were present at diferent levels in Beagles, the 100 spot 8 (identifed as double-headed protease inhibitor) being 75 0 in lower levels than in Greyhounds and the spot 26 (not 26 1 22 37 23 50 identifed) in higher levels than in the other breeds. 37 Besides these spots, multivariate PLS-DA model clustered 36 Portuguese Podengo distinctly from Beagles and Rafeiro 18 13 16 Alentejano breeds in 2-DE protein profles (Figure 5). Spots 1, 25 18,23,45,and82weretheonesthatmostcontributedtothese 46 diferences (Supplementary Figure 3). Detailed information 20 45 about MS/MS identifcation of the referred spots is presented in Table 5. 81 8 15 Inthecaseofspots45and81,theidentifcationresulted 5 in unknown proteins. However, through BLAST analysis, it was possible to observe 83% homology between the protein Figure 3: Representative dog’s saliva 2-DE gel. Spots excised for present in spot 45 and a S100 calcium binding protein A9 and digestion and identifcation by MS are numbered. 83% homology between the protein present in spot 81 and keratin 8. comparatively to the other breeds (Figure 4(a) and Sup- 4. Discussion plementary Figure 2). Te diferences between Portuguese Podengo and Beagles for nonstimulated saliva were con- In this study, the infuence of gender and acid stimulation frmedinthesalivacollectedaferacidstimulation,by on the normal dog salivary proteome of diferent breeds was univariateanalysis.Inthiscase,itwasalsopossibletoobserve studied through in-gel based proteomics approach. For all the that these breeds difer in saliva protein profle, with fve breeds, animals with a wide range of ages were included in proteins bands (E, F, I1, J, and M) observed to be diferently the study. Te number of proteins observed and identifed expressed (Table 7 and Figure 4(b)). in dog saliva through this methodology is much lower than the one reported in other studies, using LC-MS/MS 3.2.3. 2-DE Saliva Profle. 2-DE salivary protein profles [10, 27]. Nevertheless, in this study, dog gel protein profles of the several dog breeds evaluated presented diferences presented what can be of utility for studies where protein inthepercentagevolumesof7proteinspots.Trough isoforms and/or posttranslational modifcations (PTMs) are ANOVA(univariate analysis) it was observed that Portuguese of interest [11]. SDS-PAGE and 2-DE protein separations Podengo presented higher levels of 5 salivary protein spots [1 weresimultaneouslyperformedinthisstudyduetothe (� = 0.034),18(� = 0.041),22(� = 0.046),36(� = 0.036), limited amount of individual saliva. As such, SDS-PAGE and 46 (� = 0.014)], comparatively to the other breeds. was used for assessing variability and to make comparisons Among them, only spot 18 was identifed (as a light-chain using individual information. Since this approach only allows 8 BioMed Research International

Table 6: Protein bands diferently expressed (mean ± standard error) between dog breeds, in saliva collected without acid stimulation.

Bands Breed % vol � b Port. Pod. 6.07 ± .31 a b BBeagleGreyhound 9.02 ± 0.63 6.71 ± .44 0.005 b Raf. Alent. 6.86 ± .38 b Greyhound 8.16 ± 0.80 a b DPort.Pod.Raf. Alent. 12.59 ± 1.32 8.2 ± 0.46 0.005 Beagle 10.71 ± 0.48 b Port. Pod. 8.14 ± 0.53 a a,b EBeagleGreyhound 13.7 ± 1.29 11.5 ± 0.98 0.005 b Raf. Alent. 9.97 ± 0.45 a,b Greyhound 7.05 ± 0.62 a b FPort.Pod.Raf. Alent. 9.66 ± 0.75 6.39 ± 0.56 0.01 a,b Beagle 8.53 ± 0.90 a,b Greyhound 6.32 ± 0.51 a b GPort.Pod.Raf. Alent. 3.73 ± 0.72 6.35 ± 0.35 0.005 b Beagle 8.43 ± .99 b Port. Pod. 8.64 ± 1.04 a b MBeagleGreyhound 1.71 ± 0.005 7.44 ± 1.25 0.005 b Raf. Alent. 5.98 ± .83 Diferent letters mean statistically signifcant diferences between pairs, for � < 0.05.Beagle(� = 10); Portuguese Podengo (� = 7);Greyhound(� = 11); Rafeiro Alentejano (� = 13).

Scores Plot Scores Plot

15 1.5

1.0 10 %) %) 0.5 5 12.2 14.1

0.0 0 Component 2 ( Component Component 2 ( Component −0.5 −5

−1.0 −10

−20 −15 −10 −5 0 5 10 −1.5 −1.0 −0.5 0.0 0.5 1.0 1.5 2.0 Component 1 (21.8%) Component 1 (35.9%) Beagle Podengo beagle Galgo Raf alent podengo (a) (b)

Figure 4: Partial Least Square Determinant Analysis (PLS-DA) model for all dog unstimulated saliva samples SDS-PAGE bands [Δ: Portuguese Podengo (� = 7);+:Greyhound(� = 11); ◊: Rafeiro Alentejano (� = 13);andx:Beagles(� = 10)](a)andforstimulated saliva samples SDS-PAGE bands [+: Portuguese Podengo (� = 6) and Δ:Beagles(� = 8)] (b). Scaling was applied to rows when needed; � and � axes show principal component 1 (PC1) and principal component 2 (PC2), respectively, and the contribution of each of them for explaining the total variance. BioMed Research International 9

Scores Plot Scores Plot

10 20

5 10 %)

9.5 0 0

−5 Component 2 ( Component

Component 2 (18.8%) Component −10

−10

−20 −15

−15 −10 −5 0 5 10 −20 −15 −10 −5 0510 Component 1 (20.7%) Component 1 (54.2%) 1 3 1 2 4 4 (a) (b)

Figure 5: PLS-DA of dog saliva pool samples 2-DE spots of each breed (a) or considering only Portuguese Podengo and Beagles (b). Log transformation was applied to rows; � and � axis show principal component 1 (PC1) and principal component 2 (PC2), and the respective % of explanation for the total variance. 1 – Portuguese Podengo; 2- Greyhound; 3- Rafeiro Alentejano; 4- Beagle.

Table 7: Protein bands diferently expressed (mean ± standard identifed in bands and/or spots whose levels decreased with error) in saliva samples with acid stimulation between Portuguese # acid stimulation. IgG Fc-binding protein has been recently Podengo (� = 6) and Beagles (� = 8) . identifedasoneofthemoreabundantproteinsindogsaliva [10] being a protein involved in binding IgG on mucosal %vol ∗ Bands � surfaces [28]. To our knowledge, there are no other reports, Portuguese Podengo Beagle in the literature, concerning the efect of acid stimulation ± ± E7.391.64 13.21 0.54 0.003 on salivary proteome of dogs or other animals. But, our F 8.89 ± 0.57 3.98 ± 0.60 9.075e −05 results are in accordance with studies performed in humans I1 2.72 ± 1.07 10.64 ± 1.50 0.002 [12], where it was observed that acid stimulation produced J12.87± 1.21 8.40 ± 0.60 0.004 considerable major changes, namely, in proteins related to M9.12± 0.68 2.85 ± 0.60 1.6369e −05 immune function, infammation, and cell movement [12]. ∗ # Statistically signifcant diferences for � < 0.05. � of Beagles used for Also Lorenz et al. (2011) [29] observed signifcant decreases comparison was diferent that the one reported in Table 3, since for 1 animal on the relative abundance of several protein spots, in human only saliva from the collection afer stimulation contained enough amount saliva, afer citric acid stimulation. It is curious that keratin is for analysis, impeding that animal for being included in paired analysis reported in Table 3. a protein from the cytoskeleton and IgG Fc-binding protein is a gel-like component of the mucosa. Stimulation with lemon juice raised the total volume of saliva produced and, separation according to molecular masses, several proteins as such, the cotton roll needed less time in the mouth must be present in each band, making it difcult to know for getting enough saliva amounts. Such decreased time of the one (or several) responsible for changes. 2-DE profles of saliva collection, associated with fewer movements, may have saliva pools were used to add such detail. resulted in a lower incorporation of components from the No signifcant diferences among breeds or between epithelium in the samples. In fact, the possibility of variations genders were observed on total protein concentration of in the levels of these proteins being done to this efect was normal dog’s saliva. However, a decrease in the total protein recently suggested [13]. concentration afer acid stimulation was observed, especially In dogs, saliva collection without stimulation has the in males of both pure breeds Portuguese Podengo and constraint of allowing obtaining only limited volumes of Beagle. In terms of profles, proteins such as cytoskeletal saliva for performing some laboratorial techniques [30]. keratin, serum albumin, and IgGFc-binding proteins were However, if stimulation is needed it is important to have in 10 BioMed Research International mind the referred diferences in protein composition that of diferent salivary proteins. In fact, breed appears to have such stimulation is producing. even more infuence than gender. Nevertheless, that does not In the present study, we could observe that salivary mean that gender should be ignored, in dog saliva analysis. protein composition varies among diferent dog breeds, but Despite males and females presenting minimal diferences no major diferences were observed between genders. Te in salivary profles, in this study diferences in the way each reduced impact of gender in dog salivary proteome observed gender responded to stimulation were observed. is in agreement with others recently published [13]. Our From our knowledge this is one of the frst studies results go in accordance with these observations. evaluating factors afecting dog saliva electrophoretic protein Two of the breeds that most difered between them profles. More studies are needed to increase the knowledge were Portuguese Podengo and Beagle. According to Feder- about dog saliva proteome, in order to use it in research and ation Cynologique Internationale (FCI) (http://www.fci.be, diagnosis. accessed on January 31, 2018) purebred Portuguese Podengo is a primitive type of breed, hunting dog probably originating Data Availability from the ancient dogs, traditionally used for helping in rabbit or birds hunting, but without working trial [31]. Tis breed Tedatausedtosupportthefndingsofthisstudyare is also used as a watch and companion dog. Despite being available from the corresponding author upon request. a pure breed, it is expected that individuals present higher genetic variability than Beagles, since this last has been bred Acknowledgments in a controlled way, also for use in laboratory studies. Also according to FCI, purebred Beagles belong to a small-sized Tis manuscript is funded by FEDER Funds through hound group with working trial. By using clustering analysis, the Operational Programme for Competiveness Factors- to defne phylogenetic tree, this breed belongs to a cluster COMPETE and National Funds through FCT-Foundation comprised mostly by modern breeds used in hunting [32]. for Science and Technology under the Project UID/AGR/ Bands containing chains of canine serum albumin and 00115/2013 (ICAAM—University of Evora).´ Te authors IgGFc-binding protein were proteins diferently expressed acknowledge also the fnancial support from the Portuguese among dog’s breeds. One of the proteins observed to be Science Foundation (FCT) in the form of Elsa Lamy FCT present in lower amounts in Beagles, both in SDS-PAGE and investigator contract IF/01778/2013. Te Portuguese Science in 2-DE protein profles, was the full-double-headed protease Foundation (FCT) played no role in the development of inhibitor from the submandibular glands. Tis protein is the present work or upon its submission for publication. a serine type endopeptidase, which has been assumed to Asta Tvarijonaviciute was supported by the Program “Juan protect mucosal cells in mouth and oesophagus against the de la Cierva Incorporacion” of “Ministerio de Economia action of proteinases from microbial origin and/or ingested y Competitividad”, Spain, through a postdoctoral grant. with food [33]. Te authors would like to acknowledge the owners of the In the present study only a limited number of proteins three purebred kennels: Miguel Rebelo, Terra de Montoito; were observed to difer with stimulation and/or among JoseRom´ ao,˜ Casa da Torre; and Jose´ Abreu de Alpoim, breeds. Even some protein spots failed a positive identif- Alpedriche. cation, which can be related to a lower number of proteins present in curated protein databases, comparatively to other Conflicts of Interest species, such as humans. On the other hand, in this study, dogs were available from pure breed kennels and some of the Te authors declare that there are no conficts of interest diferences observed for the diferent breeds can be done to regarding the publication of this paper. diferent types of dog food consumed. Further studies, with ahighernumberofanimalsperbreed,highernumberof Supplementary Materials breedsandcontrolsfortypeoffood,andothertreatments are necessary to have a better characterization of each breed Supplementary 1. Supplementary Figure 1: PLS-DA loading saliva proteome. plots (lef) of the frst two components for analysis of SDS- PAGE bands of profles from saliva collected with and 5. Conclusions without acid stimulation in Beagles (a), Greyhound (b), and Portuguese Podengo (c); for each case, variable importance in Tis work, in line with what was hypothesized, allowed us to the projection (VIP) is presented, with 1.5 score considered as conclude that dog salivary protein composition is infuenced thresholder (right). by diferent factors. Despite the need of procedures that allow the collection of higher amounts of saliva, it is necessary Supplementary 2. Supplementary Figure 2: PLS-DA loading to be aware that techniques such as acid stimulation not plots (lef) of the frst two components for analysis of protein only induce higher salivary fow rates, but also change the bands of salivary profles from the diferent dog breeds. levels/proportion of various salivary proteins. It is also of Variable importance in the projection (VIP) is presented interest to retain that dog salivary proteome should be (right). considered according to dog breed, since this was observed Supplementary 3. Supplementary Figure 3: PLS-DA loading to be a factor responsible for variations in the proportion plots (lef) of the frst two components for analysis of protein BioMed Research International 11

spots of salivary profles from the diferent dog breeds. [14]A.J.Kobelt,P.H.Hemsworth,J.L.Barnett,andK.L.Butler, Variable importance in the projection (VIP) is presented “Sourcesofsamplingvariationinsalivacortisolindogs,” (right). Research in Veterinary Science,vol.75,no.2,pp.157–161,2003. [15] R. Bergeron, S. L. Scott, J.-P. Emond,´ F. Mercier, N. J. Cook, and A. L. Schaefer, “Physiology and behavior of dogs during air References transport,” Canadian Journal of Veterinary Research,vol.66,no. 3, pp. 211–216, 2002. [1]C.S.Chee,K.M.Chang,M.F.Loke,V.P.AngelaLoo,and V. Subrayan, “Association of potential salivary biomarkers with [16] N. A. Dreschel and D. A. Granger, “Methods of collection diabetic retinopathy and its severity in type-2 diabetes mellitus: for salivary cortisol measurement in dogs,” Hormones and a proteomic analysis by mass spectrometry,” PeerJ,vol.4,article Behavior,vol.55,no.1,pp.163–168,2009. e2022, 2016. [17]J.A.M.vanderBorg,B.Beerda,M.Ooms,A.S.deSouza,M. [2] M. Castagnola, E. Scarano, G. C. Passali et al., “Salivary vanHagen,andB.Kemp,“Evaluationofbehaviourtestingfor biomarkers and proteomics: Future diagnostic and clinical human directed aggression in dogs,” Applied Animal Behaviour utilities,” Acta Otorinorhinolaryngologica Italica,vol.37,no.2, Science,vol.128,no.1–4,pp.78–90,2010. pp. 94–101, 2017. [18] L. M. Glenk, O. D. Kothgassner, B. U. Stetina, R. Palme, [3]M.L.Cobb,K.Iskandarani,V.M.Chinchilli,andN.A. B. Kepplinger, and H. Baran, “Salivary cortisol and behavior Dreschel, “A systematic review and meta-analysis of salivary in therapy dogs during animal-assisted interventions: A pilot cortisol measurement in domestic canines,” Domestic Animal study,” Journal of Veterinary Behavior: Clinical Applications and Endocrinology,vol.57,pp.31–42,2016. Research, vol. 9, no. 3, pp. 98–106, 2014. [4] U. Jankowska, A. Latosinska, B. Skupien-Rabian, B. Swiderska, [19] P. Mongillo, E. Pitteri, P. Carnier, G. Gabai, S. Adamelli, and L. M. Dziedzicka-Wasylewska, and S. Kedracka-Krok, “Optimized Marinelli, “Does the attachment system towards owners change procedure of extraction, purifcation and proteomic analysis of in aged dogs?” Physiology & Behavior,vol.120,pp.64–69,2013. nuclear proteins from mouse brain,” Journal of Neuroscience [20]C.Pastore,F.Pirrone,F.Balzarotti,M.Faustini,L.Pierantoni, Methods,vol.261,pp.1–9,2016. and M. Albertini, “Evaluation of physiological and behavioral [5] B. Losch¨ and R. Wank, “Life-threatening angioedema of the stress-dependent parameters in agility dogs,” Journal of Veteri- tongue: the detection of the RNA of B henselae in the saliva nary Behavior: Clinical Applications and Research,vol.6,no.3, of a male patient and his dog as well as of the DNA of pp. 188–194, 2011. three Bartonella species in the blood of the patient,” BMJ Case [21]E.Lamy,G.daCosta,R.Santosetal.,“Sheepandgoat Reports,vol.2014,ArticleID203107,2014. saliva proteome analysis: A useful tool for ingestive behavior [6] P. P. Mshelbwala, A. B. Ogunkoya, and B. V. Maikai, “Detection research?” Physiology & Behavior,vol.98,no.4,pp.393–401, of rabies antigen in the saliva and brain of apparently healthy 2009. dogs slaughtered for human consumption and its public health [22] K. Jessie, O. H. Hashim, and Z. H. A. Rahim, “Protein precip- implications in Abia State, Nigeria,” ISRN Veterinary Science, itation method for salivary proteins and rehydration bufer for vol. 2013, Article ID 468043, 5 pages, 2013. two-dimensional electrophoresis,” Biotechnology,vol.7,no.4, [7]M.D.Parra,F.Tecles,S.Mart´ınez-Subiela, and J. J. Ceron,´ pp. 686–693, 2008. “C-reactive protein measurement in canine saliva,” Journal of [23] E. Lamy, C. Simoes,˜ L. Rodrigues et al., “Changes in the salivary Veterinary Diagnostic Investigation,vol.17,no.2,pp.139–144, protein profle of morbidly obese women either previously 2005. subjected to bariatric surgery or not,” Journal of Physiology and [8] C. Breitenbuecher, J. M. Belanger, K. Levy et al., “Protein Biochemistry,vol.71,no.4,pp.691–702,2015. expression and genetic variability of canine Can f 1 in golden [24] S. Lucena, M. C. Carreira, L. Rodrigues, F.Capela-Silva, A. Tvar- and Labrador retriever service dogs,” Canine Genetics and ijonaviciute, and E. Lamy, “Comparison of protein precipitation Epidemiology,vol.3,no.1,article3,2016. methods for two-dimensional electrophoresis of dog salivary [9]H.Nakanishi,T.Ohmori,M.Hara,K.Yoneyama,A.Takada, proteins,” Journal of Integrated Omics,Inpress. and K. Saito, “Identifcation of canine saliva using mRNA-based [25] L. Rodrigues, G. da Costa, C. Cordeiro, C. C. Pinheiro, F. assay,” International Journal of Legal Medicine,vol.131,no.1,pp. Amado, and E. Lamy, “Relationship between saliva protein 39–43, 2017. composition and 6-n-Propylthiouracil bitter taste responsive- [10] S. M. F.Torres, E. Furrow, C. P.Souza et al., “Salivary proteomics ness in young adults,” Journal of Sensory Studies,vol.32,no.4, of healthy dogs: an in depth catalog,” PLoS ONE,vol.13,no.1, Article ID e12275, 2017. Article ID e0191307, 2018. [26] J. Xia and D. S. Wishart, “Using MetaboAnalyst 3.0 for com- [11] D. Kang and M. H. Moon, “Development of non-gel-based prehensive metabolomics data analysis: using MetaboAnalyst two-dimensional separation of intact proteins by an on-line 3.0 for comprehensive metabolomics data analysis,” in Current hyphenation of capillary isoelectric focusing and hollow fber Protocols in Bioinformatics,A.Bateman,W.R.Pearson,L.D. fow feld-fow fractionation,” Analytical Chemistry,vol.78,no. Stein, G. D. Stormo, and J. R. Yates, Eds., pp. 14.10.1–14.10.91, 16, pp. 5789–5798, 2006. JohnWiley&Sons,Hoboken,NJ,USA,2016. [12]X.Xiao,Y.Liu,Z.Guoetal.,“Comparativeproteomicanalysis [27] P. de Sousa-Pereira, M. Cova, J. Abrantes et al., “Cross-species of the infuence of gender and acid stimulation on nor- comparison of mammalian saliva using an LC-MALDI based mal human saliva using LC/MS/MS,” PROTEOMICS—Clinical proteomic approach,” PROTEOMICS,vol.15,no.9,pp.1598– Applications, vol. 11, no. 7-8, Article ID 1600142, 2017. 1607, 2015. [13] S. Pasha, T. Inui, I. Chapple, S. Harris, L. Holcombe, and M. [28] K. Kobayashi, H. Ogata, M. Morikawa et al., “Distribution and M. Grant, “Te saliva proteome of dogs: variations within and partial characterisation of IgG Fc binding protein in various between breeds and between species,” Proteomics,vol.18,no.3- mucin producing cells and body fuids,” Gut,vol.51,no.2,pp. 4, Article ID 1700293, 2018. 169–176, 2002. 12 BioMed Research International

[29] K. Lorenz, M. Bader, A. Klaus, W. Weiss, A. Gorg,¨ and T. Hofmann, “Orosensory stimulation efects on human saliva proteome,” JournalofAgriculturalandFoodChemistry,vol.59, no. 18, pp. 10219–10231, 2011. [30] C.M.M.Lensen,C.P.H.Moons,andC.Diederich,“Salivasam- pling in dogs: How to select the most appropriate procedure for your study,” Journal of Veterinary Behavior: Clinical Applications and Research,vol.10,no.6,pp.504–512,2015. [31] C. San Jose,´ M. J. Carcel,´ M. T. Tejedor, and L. V. Monteagudo, “Microsatellite DNA markers applied to the classifcation of the Podenco Valenciano canine breed,” Italian Journal of Animal Science,vol.17,no.1,pp.49–52,2017. [32] E. A. Ostrander and R. K. Wayne, “Te canine genome,” Genome Research,vol.15,no.12,pp.1706–1716,2005. [33] H. Fritz, E. Jaumann, R. Meister, P. Pasquay, K. Hochstrasser, and E. Fink, “Proteinase inhibitors from dog submandibular glands—isolation, amino acid composition, inhibition spec- trum,” in Proceedings of the International Research Conference on Proteinas Inhibitors, pp. 257–270, Walter de Gruyter, Berlin, Germany, 1970. International Journal of Journal of Peptides Nucleic Acids

The Scientifc International Journal of International Journal of World Journal Cell Biology Microbiology Hindawi Publishing Corporation Hindawi Hindawi Hindawi Hindawi http://www.hindawi.comwww.hindawi.com Volume 20182013 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Anatomy Biochemistry Research International Research International

Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Submit your manuscripts at www.hindawi.com

Advances in Genetics Bioinformatics Research International Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Advances in International Journal of International Journal of Stem Cells BioMed Genomics Zoology International Research International Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 Virology www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018

Neuroscience Journal

Enzyme Journal of Journal of Research Parasitology Research Marine Biology Archaea Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018